Abstract Body: Introduction: Endovascular thrombectomy (EVT) efficacy and safety in large core stroke was established in multiple trials. Whether perfusion mismatch modified EVT treatment effect, and whether treatment effect is maintained with core volume≥100ml up to 24h, remained uncertain.

Methods: Patient-level data were pooled from the two trials (SELECT2 & ANGEL ASPECT) that systematically acquired perfusion imaging. Patients were stratified based on core volumes (CT-perfusion rCBF<30% RAPID software) and presence of mismatch [Mismatch ratio 1.2/mismatch volume 10ml and Mismatch ratio 1.8/mismatch volume 15ml]. EVT treatment effect estimates with tests of heterogeneity were assessed in different core volume and mismatch strata.

Results: Of 807 patients (SELECT2 352, ANGEL ASPECT 455), 762 patients had CT perfusion imaging available, with median age 67(59-74)y and NIHSS 17(14-21). Median(IQR) core (rCBF<30%) volumes were 69.5(43-95)ml (EVT 69(43-96) vs MM 70(43-94) and Tmax >6s volumes were 171(126-220)ml (EVT 168(131-219) vs MM 174(125-222)ml. Overall, EVT was associated with better functional outcomes in the subgroup with core ≥100ml (aGenOR:1.55, 95%CI 1.15-2.09, p=0.004), largely driven by those with core 100-149ml(aGenOR:1.69, 95%CI 1.15-2.48, p=0.006), rather than those with core≥150ml(aGenOR:1.28, 95%CI 0.80 -2.05, p=0.30). Furthermore, in patients with core≥70ml, EVT was associated with better clinical outcomes in subgroups with no mismatch based on both 1.2/10(n=35, aGenOR:1.96, 95%CI 1.05-3.66, p=0.035) and 1.8/15 criteria (n=152, aGenOR:1.70, 95%CI 1.24-2.33, p=0.001). Increasing CTP core volumes were associated with worse functional outcomes in both EVT (aGenOR:0.94, 95%CI 0.91-0.96, p<0.001 per 10ml) and medical management (aGenOR:0.95, 95%CI 0.93-0.97, p<0.001 per 10ml), without significant heterogeneity (p=0.35). Sensitivity analyses including additional 16 patients with MR diffusion-perfusion core(Apparent Diffusion Coefficient <620x10^-6 mm²/s) and mismatch demonstrated similar results.

Conclusions: In a pooled patient-level analysis of SELECT2 and ANGEL ASPECT, thrombectomy treatment benefit was maintained in patients with large core and without perfusion mismatch, without significant heterogeneity. Thrombectomy benefit was maintained in patients with very large core volumes. However, increased core volume was associated with worse clinical outcomes. Core volume, while not modifying treatment effect, remained strongly prognostic.